diff options
| author | Jeff Law <jlaw@ventanamicro.com> | 2024-07-31 11:30:27 -0600 |
|---|---|---|
| committer | Jeff Law <jlaw@ventanamicro.com> | 2024-07-31 11:32:30 -0600 |
| commit | 69a9ee05c68bd1fe7f5b3be86baacc8f0a599915 (patch) | |
| tree | 8d13d3f757adc39338b90d2312ed80dbc756e6eb /gcc/dwarf2codeview.cc | |
| parent | dcdefe47e18fabc340f9862a27ebd521dfa3d38c (diff) | |
| download | gcc-69a9ee05c68bd1fe7f5b3be86baacc8f0a599915.zip gcc-69a9ee05c68bd1fe7f5b3be86baacc8f0a599915.tar.gz gcc-69a9ee05c68bd1fe7f5b3be86baacc8f0a599915.tar.bz2 | |
[target/116104] Fix more rtl-checking failures in ext-dce
More enable-rtl-checking fixes for ext-dce. Very similar to the one recently
posted, this time covering more of the shift ops.
I checked all instances of CONSTANT_P guarding [U]INTVAL and fixed all that
looked wrong. I also created a dummy assembler/linker so that I could run the
GCC testsuite on gcn and verified that wasn't tripping any rtl-checking bugs in
ext-dce anymore.
Obviously this has also gone through x86 bootstrap and regression tested.
Pushing to the trunk.
pr target/116104
gcc/
* ext-dce.cc (carry_backpropagate): Change more guards of [U]INTVAL to
test CONST_INT_P rather than CONSTANT_P, fixing rtl-checking failures.
Diffstat (limited to 'gcc/dwarf2codeview.cc')
0 files changed, 0 insertions, 0 deletions